School of Engineering Department of Mechanical and Aerospace Engineering 191 Environmental Impact by Air Traffic: Assessing Aircraft Noise Nearby HK Airport Supervisor: Stephane REDONNET / MAE Student: PATKAR Aarav Vishal / AE Course: UROP 1100, Summer This project explores the use of Machine Learning techniques to uncover patterns in the relationship between aircraft noise characteristics and perceived annoyance. The dataset, sourced from prior works, includes Sound Quality Metrics (SQMs) from 46 aircraft recordings captured near Hong Kong International Airport, along with responses from 28 participants who performed pairwise comparisons of 43 recording pairs to identify the more unpleasant sound. These comparisons were used to derive an Annoyance Probability (AP) score for each recording, which served as the target variable for model training. Both continuous target and discrete target models were developed to predict the AP of each flight recording based on SQM data to determine which SQMs have the strongest influence on perceived annoyance. While Continuous target models such as Linear Regression and Polynomial Regression showed limited predictive power, their discrete counterparts, Logistic Regression and Decision Tree Classifiers, offered improved performance with average accuracy scores of 0.5624 and 0.542 respectively. Among all SQMs, only Sharpness was consistently identified as a significant predictor across both discrete models in 3 out of 4 dataset variations. Although, inconsistencies of model performance across datasets and limited sample size hindered generalization of the results. Future work will focus on refining the AP metric, model development, and exploring more advanced models or techniques to better predict annoyance based on acoustic and psycho-acoustic traits. Environmental Impact by Air Traffic: Assessing Aircraft Noise Nearby HK Airport Supervisor: Stephane REDONNET / MAE Student: SIDDIQUI Hafsa Zubair / AE Course: UROP 1100, Fall UROP 3200, Spring This project builds upon previous research on assessing the societal impact of air traffic in Hong Kong. The effect of the newly implemented three-runway-system on the Sai Kung district was investigated, following a surge in complaints from residents about the noise pollution some of them are now facing. As there is no up-to-date data available for aircraft noise levels in Sai Kung, this UROP project focused on performing field test measurements of aircraft noise in a Sai Kung residential area. Around 120 flights were recorded, their noise signals then post-processed to quantify their level of annoyance using both traditional metrics and human-centric ones. The results, currently undergoing analysis and comparisons, call for noise mitigation measures being implemented in Sai Kung.
RkJQdWJsaXNoZXIy NDk5Njg=